iran-doctor-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| DOCTOR_MCP_HTML | No | Whether to save an HTML page for every search_doctor call. The tool argument html overrides this. | true |
| DOCTOR_MCP_RETRIES | No | Number of retries per request on network errors, timeouts, HTTP 429 and 5xx responses (0–5). Other 4xx responses are never retried. | 2 |
| DOCTOR_MCP_CACHE_TTL | No | Seconds to cache identical searches per platform. Set to 0 to disable the cache. | 120 |
| DOCTOR_MCP_LOG_LEVEL | No | Logging level: debug, info, warn, error, or silent. | info (CLI: warn) |
| DOCTOR_MCP_REPORT_DIR | No | Where HTML reports are written. The default is always writable, also for global or npx installs; set this to keep reports somewhere permanent. | <OS temp>/iran-doctor-mcp/reports/ |
| DOCTOR_MCP_TIMEOUT_MS | No | Timeout of one HTTP request, in milliseconds (1000–120000). | 20000 |
| DOCTOR_MCP_REPORT_KEEP | No | Keep at most this many reports; the oldest are deleted. Set to 0 to keep all reports. | 200 |
| DOCTOR_MCP_MAX_RESPONSE_BYTES | No | Maximum response body size in bytes (default 8 MB). | 8388608 |
| DOCTOR_MCP_PLATFORM_DEADLINE_MS | No | Hard deadline for one platform's whole search, in milliseconds. A slower site is reported as failed. | 45000 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| search_doctorA | Search a doctor BY NAME across the 5 Iranian booking platforms that support name search + free appointment + price (Paziresh24 پذیرش۲۴, Doctoreto دکترتو, Nobat.ir نوبت, DrNext دکترنکست, Darmankade درمانکده). Returns a Persian report: summary, one block per doctor (profiles of the same person on different sites are merged) with first free time, price in Toman and booking links, best options, next steps and platform status. Also saves an interactive HTML page and returns its path: give that path to the user. Use Persian names (e.g. «مریم احمدی»); last name alone also works. ALWAYS pass city when the user mentions one: each platform is then searched inside that city, so local doctors are not crowded out. |
| get_doctor_detailsA | Full details for ONE doctor on ONE platform: every visit type with price, first free time, clinic address & phones (and open days for nobat.ir). The |
| list_platformsA | List the supported Iranian doctor booking platforms, what each provides, and which sites were checked but excluded (and why). |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 3 tools
The three tools occupy clearly separate roles: aggregate name search across platforms, deep-dive on a single doctor on a single platform, and static platform metadata. The descriptions explicitly state the boundary (search returns summaries/links; details requires one platform + id), so misselection is unlikely.
All three follow a consistent verb_noun snake_case pattern: search_doctor, get_doctor_details, list_platforms. Verb choice (search/get/list) also maps cleanly to the operation type.
Three tools is minimal but each earns its place for a search-aggregator domain, and the surface is not padded. It sits at the low end of the comfortable range, so an extra capability tool (e.g. search by specialty or city browse) would round it out.
Search→details→platform-info covers the discovery lifecycle, and booking links are returned, but search is name-only: there is no browse by specialty, city, or price filter, which are core user intents in doctor discovery. No booking/availability-query operation either, so the workflow dead-ends at an external link.